A RANK TEST FOR DISTINGUISHING ENVIRONMENTALLY AND GENETICALLY INDUCED DISEASE RESISTANCE IN PLANT-VARIETIES

Many species of plant pathogens vary genetically in ability to cause d isease on a given host plant. Conversely, plant populations often have variability with respect to a gene or series of genes that provide re sistance to a given pathogen (Leonard, 1987, in Populations of Plant P athogens: Their Dynamics and Genetics, 163-181; Christ, Person, and Po pe, 1987, in Populations of Plant Pathogens: Their Dynamics and Geneti cs, 7-21). These dynamics have become apparent, as there have been man y cases in which crop plants bred for homogeneous high-level genetic r esistance to a disease have lost the resistance, sometimes catastrophi cally (Day and Wolfe, 1987, in Populations of Plant Pathogens: Their D ynamics and Genetics, 3-7). This has led plant protectionists to seek ways to develop stable resistance to plant pathogens that are variable by nature. A nonparametric procedure that tests for variation among p opulations of a specific pathogen at different-locations is described. A randomisation approach is used to compute critical values and p-val ues. A small simulation study is carried out to examine the performanc e of the test. The same procedure can be used to determine the reactio n of a specific crop variety to a pathogen under different environment al conditions or in different locations. The procedure is demonstrated using data on the fungus Colletotrichum graminicola (Ces.) Wilson, wh ich is a pathogen on the cereal crop Sorghum bicolor (Moench). Disease reactions of genetically different sorghum varieties to specific stra ins of the fungus in different environments are recorded. From these, varieties are selected that do not exhibit interaction with environmen t. These varieties are then used to identify which. locations have dis tinct pathogen populations. Once a set of locations has been identifi ed as having distinct pathogen populations, the breeding materials are planted and selected for insignificant genotype by location (pathogen population) interaction, i.e., for genetic stability with respect to a given disease.